Well, I can wait for 19.07
Has my eyesight worsened, or is the "config.seed
" file missing at "http://downloads.openwrt.org/releases/18.06.5/targets/mvebu/cortexa9/"?
It seems to be missing from 18.06.5 entirely.
BTW: There seems to be a new package for WRT1900ACS in 19.07-rc1:
But i cant find any changlog in kaloz repo - but here:
The driver seems to be under development ... am i right?
Regards, Andreas
Regards, Andreas
What is the practical usage of it?
It is pretty much just the defaults and "build all packages":
target defaults + luci + all_kmods + all_nonshared (+a few options for flagging a buildbot build and SDK & IB)
Yeah, I guess that you would prefer to use the exact same config for the driver build, but I guess you should be able to reach the same kernel options easily.
-
This driver is not being developed any more. No new features, not even bugfixes. Right now, this driver is solely being maintained so that it maintains compatibility with the latest OpenWrt version.
-
Marvell's entire wireless division has been bought and sold to NXP. Belkin, the parent company of Linksys, has been bought by Foxconn. This leaves little chance for future updates/development of the WRT AC Devices in terms of their wireless capability (which was never upstreamed to the official Linux kernel). If it was, it may have been still maintained.
-
The patch that will make the kaloz repo match OpenWrt's version is this: https://github.com/kaloz/mwlwifi/pull/369. As you can see, no one with maintainer's rights is able to merge this patch into the official one. Kaloz's repo is now officially behind OpenWrt's version (at least for master and 19.07).
No, I do not prefer to use the same options, I have to use the exact same options, or the packages will not be compatible with the installed kernel.
I asked the developers about the missing file, and somebody answered that can use a configuration from the previous version and change a couple of lines. I will try this during the weekend.
Maintainer said it only affects regulatory rules
This fork uses 7.3.0.21. It works fine on by WRT1900ACS(v1).
Built it against master, but WiFi leds stopped working. Have yet to run an iPerf test.
Differences are that it uses SWBA (software beacon alert(?) or software block ack(?)).
EDIT:
Results from a simple iPerf2 test WiFi -> gigabit ethernet. 6 clients connected at the same time. (3 of which were active incl. the test client). Looks like it has a ramp up behavior, but it sustains +700Mbits/sec throughput really well.
iperf -c 192.168.2.123 -t 60 -i 1
Client connecting to 192.168.2.123, TCP port 5001
TCP window size: 129 KByte (default)
------------------------------------------------------------
[ 4] local 192.168.2.108 port 54926 connected with 192.168.2.123 port 5001
[ ID] Interval Transfer Bandwidth
[ 4] 0.0- 1.0 sec 128 KBytes 1.05 Mbits/sec
[ 4] 1.0- 2.0 sec 512 KBytes 4.19 Mbits/sec
[ 4] 2.0- 3.0 sec 512 KBytes 4.19 Mbits/sec
[ 4] 3.0- 4.0 sec 768 KBytes 6.29 Mbits/sec
[ 4] 4.0- 5.0 sec 1.25 MBytes 10.5 Mbits/sec
[ 4] 5.0- 6.0 sec 2.12 MBytes 17.8 Mbits/sec
[ 4] 6.0- 7.0 sec 3.75 MBytes 31.5 Mbits/sec
[ 4] 7.0- 8.0 sec 11.2 MBytes 94.4 Mbits/sec
[ 4] 8.0- 9.0 sec 16.0 MBytes 134 Mbits/sec
[ 4] 9.0-10.0 sec 27.6 MBytes 232 Mbits/sec
[ 4] 10.0-11.0 sec 36.8 MBytes 308 Mbits/sec
[ 4] 11.0-12.0 sec 48.2 MBytes 405 Mbits/sec
[ 4] 12.0-13.0 sec 58.0 MBytes 487 Mbits/sec
[ 4] 13.0-14.0 sec 66.0 MBytes 554 Mbits/sec
[ 4] 14.0-15.0 sec 74.6 MBytes 626 Mbits/sec
[ 4] 15.0-16.0 sec 76.0 MBytes 638 Mbits/sec
[ 4] 16.0-17.0 sec 69.9 MBytes 586 Mbits/sec
[ 4] 17.0-18.0 sec 73.4 MBytes 616 Mbits/sec
[ 4] 18.0-19.0 sec 69.8 MBytes 585 Mbits/sec
[ 4] 19.0-20.0 sec 73.4 MBytes 616 Mbits/sec
[ 4] 20.0-21.0 sec 76.6 MBytes 643 Mbits/sec
[ 4] 21.0-22.0 sec 78.2 MBytes 656 Mbits/sec
[ 4] 22.0-23.0 sec 78.6 MBytes 660 Mbits/sec
[ 4] 23.0-24.0 sec 78.8 MBytes 661 Mbits/sec
[ 4] 24.0-25.0 sec 79.4 MBytes 666 Mbits/sec
[ 4] 25.0-26.0 sec 79.1 MBytes 664 Mbits/sec
[ 4] 26.0-27.0 sec 85.6 MBytes 718 Mbits/sec
[ 4] 27.0-28.0 sec 85.6 MBytes 718 Mbits/sec
[ 4] 28.0-29.0 sec 88.5 MBytes 742 Mbits/sec
[ 4] 29.0-30.0 sec 87.0 MBytes 730 Mbits/sec
[ 4] 30.0-31.0 sec 85.1 MBytes 714 Mbits/sec
[ 4] 31.0-32.0 sec 88.0 MBytes 738 Mbits/sec
[ 4] 32.0-33.0 sec 85.6 MBytes 718 Mbits/sec
[ 4] 33.0-34.0 sec 86.1 MBytes 722 Mbits/sec
[ 4] 34.0-35.0 sec 88.5 MBytes 742 Mbits/sec
[ 4] 35.0-36.0 sec 87.0 MBytes 730 Mbits/sec
[ 4] 36.0-37.0 sec 87.2 MBytes 732 Mbits/sec
[ 4] 37.0-38.0 sec 88.0 MBytes 738 Mbits/sec
[ 4] 38.0-39.0 sec 90.2 MBytes 757 Mbits/sec
[ 4] 39.0-40.0 sec 87.1 MBytes 731 Mbits/sec
[ 4] 40.0-41.0 sec 88.8 MBytes 744 Mbits/sec
[ 4] 41.0-42.0 sec 89.8 MBytes 753 Mbits/sec
[ 4] 42.0-43.0 sec 90.6 MBytes 760 Mbits/sec
[ 4] 43.0-44.0 sec 88.9 MBytes 746 Mbits/sec
[ 4] 44.0-45.0 sec 90.6 MBytes 760 Mbits/sec
[ 4] 45.0-46.0 sec 85.6 MBytes 718 Mbits/sec
[ 4] 46.0-47.0 sec 86.2 MBytes 724 Mbits/sec
[ 4] 47.0-48.0 sec 87.0 MBytes 730 Mbits/sec
[ 4] 48.0-49.0 sec 87.2 MBytes 732 Mbits/sec
[ 4] 49.0-50.0 sec 88.2 MBytes 740 Mbits/sec
[ 4] 50.0-51.0 sec 89.4 MBytes 750 Mbits/sec
[ 4] 51.0-52.0 sec 88.1 MBytes 739 Mbits/sec
[ 4] 52.0-53.0 sec 89.6 MBytes 752 Mbits/sec
[ 4] 53.0-54.0 sec 88.8 MBytes 744 Mbits/sec
[ 4] 54.0-55.0 sec 89.2 MBytes 749 Mbits/sec
[ 4] 55.0-56.0 sec 88.2 MBytes 740 Mbits/sec
[ 4] 56.0-57.0 sec 90.1 MBytes 756 Mbits/sec
[ 4] 57.0-58.0 sec 88.6 MBytes 743 Mbits/sec
[ 4] 58.0-59.0 sec 88.9 MBytes 746 Mbits/sec
[ 4] 59.0-60.0 sec 84.9 MBytes 712 Mbits/sec
[ 4] 0.0-60.1 sec 4.07 GBytes 582 Mbits/sec
Nice results there.
Unfortunate
I rebased the marvell repo against the kaloz repo. These commits are what I got:
commit 1a54a1e42ec476a47e717ee127841824d331f6a6 (HEAD)
Author: panda-mute <wxuzju@gmail.com>
Date: Fri Mar 30 16:43:16 2018 +0800
Revert "Add missing "int (*mcast_cts)(struct ieee80211_hw *hw, bool enable)""
This reverts commit 9baa8bdae352b24cf52d96e02e21b26e4218e3a6.
commit cc4b39b2a32d7a363ec086ededfeef5c7b03a814
Author: panda-mute <wxuzju@gmail.com>
Date: Fri Mar 30 14:39:22 2018 +0800
use utils_tid_to_ac instead of mwl_tx_tid_queue_mapping
commit 78dbff5633a06c1a9f5ba388daafa89964818822
Author: panda-mute <wxuzju@gmail.com>
Date: Thu Mar 29 13:51:50 2018 +0800
add missing function mwl_tx_tid_queue_mapping
commit b444d4072111f4c53662a735960c44ba1bbdda44
Author: panda-mute <wxuzju@gmail.com>
Date: Thu Nov 16 10:27:58 2017 +0800
fix invalid multicast rate issue
commit a87b51fd8a3930ca27d56345f15f4e8b25923beb
Author: panda-mute <wxuzju@gmail.com>
Date: Wed Nov 15 10:08:20 2017 +0800
Change multicast rate from 1Mbps to 54Mbps
commit 4b75de0aad10091e758ea78be726de2f2a327931
Author: panda-mute <wxuzju@gmail.com>
Date: Wed Sep 20 23:22:33 2017 +0800
Added code to support software BA streams.
commit 3a1a957f54837be2abbb5b9874ea5915e4206804
Author: panda-mute <wxuzju@gmail.com>
Date: Wed Sep 20 23:06:45 2017 +0800
Upgrade 88W8864 firmware to 7.3.0.21
No idea if there's anything useful here.
edit: I did the opposite. No extra commits. Meaning the marvell wireless repo has more stuff.
Revert "Add missing "int (*mcast_cts)(struct ieee80211_hw *hw, bool enable)""
This was reverted because it was upstreamed.
use utils_tid_to_ac instead of mwl_tx_tid_queue_mapping
add missing function mwl_tx_tid_queue_mapping
Same as above.
fix invalid multicast rate issue
Change multicast rate from 1Mbps to 54Mbps
I think I reverted this. You must live/work in an open home/office for this to generate any benefits. The higher the rate, the lower the distance a client can be from the AP to still maintain a connection.
Added code to support software BA streams.
This is either Software Block Acknowledgement or Software Beacon Alerts, I don't think we can ever be sure.
We need to run more testing to be sure to switch to this repo. At the moment, I think I'm getting sudden dropouts (but not disconnects) - I think the data path is being messed up.
Is there a Marvel repo? Where is it?
I'm not sure but shouldn't multicast rate = lowest basic rate?
https://tools.ietf.org/id/draft-ietf-mboned-ieee802-mcast-problems-01.html
basic rate
The "lowest common denominator" data rate at which multicast and broadcast traffic is generally transmitted.
I think, that is just a clone og kaloz Repo with no further development: